Thanks for the comments so far. It does sound like I will not be
overlapping with other peoples work very much.
A minor clarification on what I'm also planning on doing. I did
mention that it would probably have compile problems currently on
other platforms. This is probably not true since no one uses any
USB specific ioctl() calls that could fail. Right now its all
read()/write()'s. In that regard I was thinking more along the lines
of future USB drivers that could kill compiles if they accessed
ioctl()'s.
Also, at least on Linux most of the usb work is done by scanner.o.
Therefore I'd like to make a generic wrapper in case other platforms
don't do as much work. Also, I'm thinking of moving the current USB
code from the snapscan driver that simulates sanei_scsi_cmd()
function calls and turns them into USB calls. Since current USB
devices are usually SCSI over USB it should be handy (can be shared
between the epson and snapscan today even).
This part is lower on my feature wishlist. I'll be working on the
config files and a way to make SCSI and USB scanners coexist happily
at first.
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com
This archive was generated by hypermail 2b29 : Wed Oct 11 2000 - 11:25:40 PDT